Risk Factor Analysis of the Development of New Neurological Deficits Following Supplementary Motor Area Resection

See the corresponding editorial in this issue, pp 4–6. J Neurosurg 119:7–14, 2013 ©AANS, 2013

Risk factor analysis of the development of new neurological deficits following supplementary motor area resection

Clinical article

Young-Hoon Kim, M.D.,1,2 Chi Heon Kim, M.D., Ph.D.,1,3 June Sic Kim, Ph.D.,1,3 Sang Kun Lee, M.D., Ph.D.,4 Jung Ho Han, M.D.,1,2 Chae-Yong Kim, M.D., Ph.D.,1,2 and Chun Kee Chung, M.D., Ph.D.1,3 1Department of Neurosurgery, Seoul National University College of Medicine; Departments of 3Neurosurgery and 4Neurology, Seoul National University Hospital, Seoul; and 2Department of Neurosurgery, Seoul National University Bundang Hospital, Seongnam-si, Gyeonggi-do, Korea

Object. Supplementary motor area (SMA) resection often induces postoperative contralateral hemiparesis or speech disturbance. This study was performed to assess the neurological impairments that often follow SMA resection and to assess the risk factors associated with these postoperative deficits. Methods. The records for patients who had undergone SMA resection for pharmacologically intractable epilepsy between 1994 and 2010 were gleaned from an epilepsy surgery database and retrospectively reviewed in this study. Results. Forty-three patients with pharmacologically intractable epilepsy underwent SMA resection with intraoperative cortical stimulation and mapping while under awake anesthesia. The mean patient age was 31.7 years (range 15–63 years), and the mean duration and frequency of seizures were 10.4 years (range 0.1–30 years) and 14.6 per month (range 0.1–150 per month), respectively. Pathological examination of the brain revealed cortical dysplasia in 18 patients (41.9%), tumors in 16 patients (37.2%), and other lesions in 9 patients (20.9%). The mean duration of the follow-up period was 84.0 months (range 24–169 months). After SMA resection, 23 patients (53.5%) experienced neurological deficits. Three patients (7.0%) experienced permanent deficits, and 20 (46.5%) experienced symptoms that were transient. All permanent deficits involved contralateral weakness, whereas the transient symptoms patients experienced were varied, including contralateral weaknesses in 15, apraxia in 1, sensory disturbances in 1, and dysphasia in 6. Thirteen patients recovered completely within 1 month. Univariate analysis revealed that resection of the SMA proper, a shorter lifetime seizure history (< 10 years), and resection of the cingulate gyrus in addition to the SMA were associated with the development of neurological deficits (p = 0.078, 0.069, and 0.023, respectively). Cingulate gyrus resection was the only risk factor identified on multivariate analysis (p = 0.027, OR 6.530, 95% CI 1.234–34.562). Conclusions. Resection of the cingulate gyrus in addition to the SMA was significantly associated with the development of postoperative neurological impairment. (http://thejns.org/doi/abs/10.3171/2013.3.JNS121492)

Key Words • supplementary motor area • cingulate gyrus • resection • neurological deficit • epilepsy

he supplementary motor area, which is situated on cal impairments following unilateral SMA resection, in- the medial aspect of the superior frontal gyrus, is cluding transient contralateral hemiparesis, hemiplegia, mainly associated with motor function and has re- and speech deficits. Since that first report, several other Tciprocal connections with multiple motor areas, includ- studies have confirmed these deficits and revealed that ing the primary motor cortex, prefrontal cortex, cingulate they are features of SMA syndrome.1,2,4–7,9,10,17–20,23,26 gyri, basal ganglia, and contralateral SMA.12 In 1951, These postoperative motor and speech disorders gen- Penfield and Welch16 identified postoperative neurologi- erally resolve within several weeks or months without severe neurological sequelae.1,11,19 Several compensatory mechanisms have been suggested as being responsible Abbreviations used in this paper: EEG = electroencephalography; 9,18 MCC = midcingulate cortex; pACC = perigenual anterior cingu- for recovery from SMA syndrome. Thus, previous late cortex; PCC = posterior cingulate cortex; RSC = retrosplenial studies have focused more on the characteristics of SMA cortex; SMA = supplementary motor area; VCA = vertical plane syndrome, which are mild and transient, as well as the through anterior commissure. mechanism by which they resolve.18 However, according

J Neurosurg / Volume 119 / July 2013 7 Y. H. Kim et al. to recent reports, not all patients who undergo SMA re- status of each patient were assessed at regular clinical fol- section experience postoperative neurological deficits.18,20 low-ups, which occurred 1 month after surgery and sub- Patients with motor weakness or speech impairments sequently every 2 or 3 months. Each assessment was com- may face greater rehabilitative difficulties when completed by the same neurologist and neurosurgeon. Seizure pared with patients who do not experience these deficits. outcome was assessed according to the Engel classifica- In addition, few studies have reported on factors associ- tion. Postoperative motor weakness was described ac- ated with postoperative neurological deficits commonly cording to a muscle strength grading score: Grade 0, no experienced after SMA resection or why SMA syndrome contraction; Grade 1, flicker or trace contraction; Grade 2, occurs in only a subset of patients. active movement with gravity eliminated; Grade 3, active Therefore, this study was performed to describe post- movement through full range of motion against gravity; operative neurological impairments that follow SMA re- Grade 4, active movement against resistance; and Grade section and to assess the risk factors for these deficits. 5, normal strength against full resistance. Postoperative MRI was completed 3 months after surgery to measure Methods the resected area. The resected area of the cingulate gyrus was defined as the area between the cingulate sulcus Study Design and the callosal sulcus of the corpus callosum, which was The institutional review board of the Seoul National surgically removed and visualized on sagittal postopera- University Hospital approved this study and exemption tive MR images. of informed consent because the study was retrospective. Several demographic and clinical factors, including The records for patients who had undergone SMA resec- patient sex, age, seizure duration, seizure frequency, sei- tion for pharmacologically intractable epilepsy between zure type, presence of predisposing factors, side of surgery, 1994 and 2010 were gleaned from the epilepsy surgery surgery type, resected area, whether cingulate gyrus resec- database and retrospectively reviewed in this study. In all tion was performed, amount of time required for resection, cases, the SMA was limited posteriorly by the precentral histological type, and seizure outcome, were evaluated via sulcus, inferiorly by the cingulate sulcus, anteriorly by the risk factor analysis for the development of postoperative most rostral point of the genu of the corpus callosum, and neurological deficits following SMA resection. laterally by the superior frontal sulcus according to previ- ously described methods.7,14,16 The rostral pre-SMA was Statistical Analysis defined as the anterior part of the SMA to the line vertical All data are presented as the means ± standard de- to the anterior commissure–posterior commissure plane viation in addition to the range. Risk factors for postop- and crossing the anterior commissure, or the VCA line. erative neurological deficits were analyzed using logistic The caudal SMA proper was defined as the posterior part 18,27 regression analysis. To reduce the risk of Type II errors of the SMA to the VCA line, as per previous reports. due to modest sample size, variables were considered for Awake Surgery With Intraoperative Monitoring multivariate analysis only if they were associated with a dependent variable in each analysis that was significant at Preoperative evaluation to localize an ictal onset zone a level of p < 0.150. Values of p < 0.050 were considered included a routine history, neurological examination, in- statistically significant. All statistical analyses were per- terictal EEG, ictal video-EEG monitoring (for at least 3 formed using SPSS (version 17.0.1, SPSS, Inc.). typical seizures), brain MRI, PET, and interictal and ictal SPECT. While patients were under general anesthesia, Results subdural electrodes were implanted near the presumed ictal onset zone according to video-EEG monitoring results Forty-three patients with pharmacologically intracta- if no definite lesion was identified via preoperative MRI. ble epilepsy underwent SMA resection with intraoperative Resection was performed in patients under awake an- cortical stimulation and mapping while under awake anes- esthesia by using intraoperative cortical mapping and mon- thesia. Twenty-five patients were male and 18 were female. itoring. The detailed operative methods are described in The mean patient age was 31.7 ± 11.2 years (range 15–63 our previous report.8 Standard cortical mapping was per- years). The mean duration of the seizure history and sei- formed using an Ojemann stimulator, which is a constant- zure frequency were 10.4 ± 8.1 years (range 0.1–30.0 years) current generator that produces a train of biphasic square and 14.6 ± 29.0/month (range 0.1–150.0/month), respective- wave pulses (at a rate of 60 Hz, 1 msec/phase) to minimize ly. Tumor-like lesions were found in 18 patients (41.9%) and the possibility of inducing a seizure. To locate the primary cortical lesions were found in 9 (20.9%) via preoperative motor cortex, stimuli were applied in 1-mA increments, MRI. Sixteen patients (37.2%) showed no definite lesions starting at 1 mA up to a maximum of 10 mA. An epileptol- on MRI. The radiological characteristics of all patients are ogist attended the entire process and monitored the patient summarized in Table 1. during cortical stimulation and resection. Surgical removal continued until either the target epileptogenic zone was Surgical Outcome totally removed or the onset of unexpected neurological Twenty-four patients (55.8%) underwent subdural grid deficits occurred. implantation to localize the ictal onset zone before resection. The resective surgeries included 18 lesionectomies Postoperative Evaluation (41.9%), 13 corticectomies (30.2%), and 12 frontal lobecto- Postoperative seizure outcome and the neurological mies (27.9%). Resection of the SMA proper was performed

8 J Neurosurg / Volume 119 / July 2013 Neurological deficits after SMA resection

TABLE 1: Clinicoradiological characteristics of 43 patients who TABLE 2: Diagnostic and surgical results of 43 patients who underwent SMA resection* underwent SMA resection*

Parameter Value Parameter Value patient sex (%) type of surgery (%) M 25 (58.1) lesionectomy 18 (41.9) F 18 (41.9) corticectomy 13 (30.2) mean age in yrs (range) 31.7 ± 11.2 (15–63) frontal lobectomy 12 (27.9) type of Sz (%) side of surgery (%) simple partial Sz 7 (16.3) rt 19 (44.2) CPSs 6 (14.0) lt 24 (55.8) CPSs w/ secondary generalization 15 (34.9) resected area (%) generalized tonic-clonic Sz 14 (32.6) SMA proper 17 (39.5) other 1 (2.3) pre-SMA 11 (25.6) mean duration of Szs in yrs (range) 10.4 ± 8.1 (0.1–30.0) SMA proper & pre-SMA 15 (34.9) mean frequency of Szs/mo (range) 14.6 ± 29.0 (0.1–150.0) cingulate gyrus resection (%) 15 (34.9) predisposing factors for Sz (%) mean op time in mins (range) 300.0 ± 70.3 (130–425) trauma 9 (20.9) mean hospital stay in days (range) 7.2 ± 3.8 (3–24) CNS infection 2 (4.7) pathological exam (%) febrile convulsion 4 (9.3) astrocytic tumor 4 (9.3) developmental anomaly 2 (4.7) oligodendrocytic tumor 10 (23.3) no 26 (60.5) other tumor 2 (4.7) no. of preop AEDs (%) cortical dysplasia 18 (41.9) 1 11 (25.6) other 9 (20.9) 2 18 (41.9) postop neurological deficit (%) ≥3 14 (32.6) permanent 3 (7.0) preop MRI findings (%) transient 20 (46.5) tumor-like lesion 18 (41.9) no 20 (46.5) cortical lesion 9 (20.9) Engel class at last FU (%) no lesion 16 (37.2) I 21 (48.8) pre-resective subdural grid implantation (%) 24 (55.8) II 9 (20.9) III 6 (14.0) * Values indicate the number of patients, unless indicated otherwise. IV 7 (16.3) Mean values are presented ± SD. Abbreviations: AED = antiepileptic drug; CPS = complex partial seizure; Sz = seizure. mean clinical FU duration in mos (range) 84.0 ± 49.0 (24–169) * Values indicate the number of patients, unless indicated otherwise. in 17 patients (39.5%), pre-SMA resection was performed Mean values are presented ± SD. Abbreviation: FU = follow-up. in 11 patients (25.6%), and resection of both areas was performed in 15 patients (34.9%). Additional cingulate gyrus permanent deficits, and 20 had transient deficits (46.5%). resection was performed in 15 patients (34.9%). Histo- Permanent deficits corresponded to contralateral Grade 4 pathological examinations revealed 4 astrocytic tumors, 10 motor weakness. One patient (Case 5) underwent resec- oligodendrocytic tumors, 2 other tumors, 18 cortical dys- tion of the most posterior part of the superior, middle, and plasias, and 9 other lesions. inferior frontal gyri and experienced permanent Grade 4 At the last follow-up evaluation for seizure outcome, hand weakness. A second patient (Case 12) demonstrated 21 patients (48.8%) had Engel Class I (seizure free), 9 pa- Grade 4 leg weakness after resection of the SMA proper. tients (20.9%) had Class II (rare seizures), and 6 (14.0%) Sudden hand weakness occurred during resection of the had Class III (worthwhile improvement). The seizure most posterior area of the SMA in the third patient (Case control rate was 83.7% (Engel Classes I–III). The mean 25); therefore, the procedure was stopped immediately. time elapsed from surgery to the final clinical follow-up The patient in this case experienced permanent Grade 4 was 84.0 ± 49.0 months (range 24–169 months). The sur- upper- and lower-extremity weakness. Transient deficits gical results are summarized in Table 2. included contralateral motor weakness in 15 patients, sensory disturbances in 1 patient, dyspraxia in 1 patient, Postoperative Neurological Deficits and motor dysphasia or dysarthria in 6 patients. Three Of the 43 patients with SMA resection, 23 patients patients experienced both motor weakness and dysphasia, (53.5%) experienced new postoperative neurological defi- and 1 patient experienced both weakness and dyspraxia. cits (Table 3). Three of these patients (7.0%) experienced Dyspraxia was characterized by impairment of initiation

J Neurosurg / Volume 119 / July 2013 9 Y. H. Kim et al.

TABLE 3: Clinical findings in 43 patients who underwent SMA resection*

Case Age (yrs), Recovery Resected Cingulate Gyrus Engel No. Sex Neurological Deficit (days) Area Resection Dx Class 1 19, F — — SMA proper no CH II 2 57, F motor dysphasia 30 SMA proper yes CA I 3 39, M hand weakness (Gr 4) 270 SMA proper yes AO I 4 31, M — — SMA proper no CD II 5 36, M hand weakness (Gr 4) permanent whole SMA† yes CD I 6 34, M motor dysphasia, UE & LE weakness (Gr 4) 150 whole SMA yes ODG I 7 15, F UE & LE weakness (Gr 4) 42 whole SMA yes CD I 8 20, M — — pre-SMA yes CS II 9 25, M — — SMA proper no CD IV 10 27, F UE weakness (Gr 3) 3 SMA proper yes CC III 11 31, M — — pre-SMA no CS I 12 43, F LE weakness (Gr 4) permanent SMA proper no Ast IV 13 43, F UE & LE weakness (Gr 4) 60 whole SMA yes ODG I 14 24, M — — whole SMA yes CD II 15 24, M UE & LE weakness (Gr 4) 14 SMA proper no GGO I 16 26, M UE & LE weakness (Gr 3) 3 whole SMA no CD I 17 28, F motor dysphasia 30 Pre-SMA no CD III 18 23, M — — whole SMA no CD IV 19 22, F — — pre-SMA no AVM I 20 23, F UE & LE weakness (Gr 4) 5 SMA proper yes CI I 21 28, M motor dysphasia 30 whole SMA no CD III 22 17, F UE weakness (Gr 3) 30 pre-SMA no CD II 23 23, M — — whole SMA yes CD II 24 45, M UE & LE weakness (Gr 2) 30 whole SMA no ODG IV 25 33, F UE & LE weakness (Gr 4) permanent SMA proper‡ no GBL III 26 28, F — — pre-SMA no CM IV 27 19, M — — pre-SMA no CD I 28 37, F — — pre-SMA no CD I 29 24, M — — SMA proper no ODG I 30 40, M motor dysphasia, UE weakness (Gr 4) 6 SMA proper yes ODG II 31 32, F — — SMA proper no ODG I 32 51, F — — pre-SMA no CD I 33 30, M UE weakness (Gr 3) 60 whole SMA no CD III 34 40, M hand sensory disturbance 1500 whole SMA yes ODG I 35 30, F UE & LE weakness (Gr 4), motor dyspraxia 8 whole SMA yes CD I 36 19, M — — SMA proper no RG IV 37 16, M — — whole SMA no CD II 38 38, M LE weakness (Gr 4) 42 SMA proper no AA III 39 63, M motor dysphasia, ankle weakness (Gr 4) 7 SMA proper no AO IV 40 30, M LE weakness (Gr 3) 1 pre-SMA yes CD I 41 52, M — — whole SMA no ODG II 42 37, F — — SMA proper no EVN I 43 41, F — — pre-SMA no AA I

* AA = anaplastic astrocytoma; AO = anaplastic oligodendroglioma; Ast = astrocytoma; AVM = arteriovenous malformation; CA = cavernous angioma; CC = cortical contusion; CD = cortical dysplasia; CH = cortical heterotopia; CI = cortical ischemia; CM = cerebromalacia; CS = cortical scar; Dx = his- topathological diagnosis; EVN = extraventricular neurocytoma; GBL = glioblastoma; GGO = ganglioglioma; Gr = grade; LE = lower extremity; ODG = oligodendroglioma; RG = reactive gliosis; UE = upper extremity; whole SMA = whole SMA, including SMA proper and pre-SMA. † The patient underwent resection of the most posterior part of the superior, middle, and inferior frontal gyri. ‡ The patient experienced development of hand weakness during surgery, and the resection stopped.

10 J Neurosurg / Volume 119 / July 2013 Neurological deficits after SMA resection and slowing of skilled and complex movements such as Risk Factors for Postoperative Neurological Deficits putting on and removing clothing. Of the 20 patients with Associations between postoperative neurological defi- transient deficits, 13 patients (65%) fully recovered within ± cits and several clinicoanatomical factors were evaluated 1 month of surgery. The mean recovery time was 116 (Table 4). Univariate analysis revealed that cingulate gyrus 332 days (range 1–1500 days). resection, shorter lifetime history of seizure (< 10 years), All 3 patients with permanent hemiparesis underwent and SMA proper resection (p = 0.023, 0.069, 0.078, respec- resection of the most posterior part of the SMA proper, tively) were associated with postoperative neurological im- which is just adjacent to the precentral sulcus. Resection pairments after SMA resection. However, cingulate gyrus of the most posterior area of the SMA just anterior to the resection (p = 0.027, OR 6.530, 95% CI 1.234–34.562) was precentral sulcus was performed in 9 patients. Of these the only significant risk factor for postoperative neurologi- 9 patients, 7 (77.8%) experienced postoperative motor cal deficits identified via multivariate analysis. weakness. Of the remaining 34 patients, 16 (47%) experienced postoperative neurological deficits (p = 0.142). Moreover, the permanent neurological risk in the patients Illustrative Case with the most posterior SMA resection was significantly higher compared with that in the remaining patients (p A 30-year-old woman (Case 35) with pharmacologi- = 0.007). However, cingulate gyrus resection and patho- cally intractable seizures was seen at our hospital. Al- logical diagnosis were not associated with permanent though she was treated with valproate, carbamazepine, neurological deficits (p = 1.000 and 0.545, respectively). and topiramate, both separately and in combination, she The risk of postoperative neurological deficits was experienced several seizure attacks per day for 23 years. higher in the patients with SMA proper resection (20 Her seizures initially involved extension of her left arm [62.5%] of 32) than in those with only pre-SMA resection and evolution into generalized tonic-clonic seizures with- (3 [27.3%] of 11; p = 0.078). We analyzed the association in a few minutes. Ictal video-EEG findings suggested a between postoperative neurological deficits and resected frontotemporal seizure focus; however, no specific lesion area, which was stratified by concurrent cingulate gyrus was found on 3.0-T MRI, and no metabolic abnormality resection (Fig. 1). Overall, cingulate gyrus resection was was detected on PET. significantly related to postoperative neurological deficits Subdural grids were implanted for localization of an (12 [80.0%] of 15 vs 11 [39.3%] of 28; p = 0.023). Resec- ictal onset zone (Fig. 2A). Invasive monitoring revealed tion of the SMA proper along with additional removal of that the seizure activity initiated at both the right SMA the cingulate gyrus was most strongly related to postop- and the cingulate gyrus. Seven days after the first crani- erative neurological deficits as compared with any other otomy, she underwent corticectomy of the right SMA and type of resection (11 [84.6%] of 13; p = 0.009). Resection cingulate gyrus while under local anesthesia (Fig. 2B). of the pre-SMA without removal of the cingulate gyrus She experienced Grade 4 weakness of her left side and was the safest; however, this finding did not reach statis- mild motor dyspraxia; however, she was fully recovered tical significance (2 [22.2%] of 9; p = 0.059). Stratified 8 days after the second surgery. Postoperative MRI vi- analysis revealed that resection of the SMA proper and sualized the resected area (Fig. 2C–E). Postoperatively, the cingulate gyrus was related to postoperative neuro- she remained seizure free without using any antiepileptic logical impairments. drugs, and she had no neurological symptoms for 4 years.

Discussion Not all patients who undergo SMA resection experience postoperative neurological impairments such as SMA syndrome. Authors of initial studies reported that the SMA syndrome occurred in every patient who underwent SMA resection.11,19 However, the overall incidence rate of SMA syndrome is variable, ranging from 23% to 100% according to recent series.5,7,18,20 In the present study, only 53.5% of patients experienced postoperative neurological deficits. In other words, while some patients who undergo SMA resection recover well without any neurological complications, others experience motor weakness or language problems lasting as long as several months. Moreover, these neurological conditions greatly impact a patient’s quality of life. Therefore, it is essential to explore the risk factors associated with neurological Fig. 1. Analysis of the association between the development of new postoperative neurological deficits and the resected area stratified by deficits that follow SMA resection. Controlling these risk additional cingulate gyrus resection in patients who underwent SMA factors before and during resection may lead to improved resection. Stratified analysis revealed that cingulate gyrus resection clinical care. together with resection of the SMA proper was associated with postop- Resection of the cingulate gyrus was the most nota- erative neurological deficits. ble risk factor for postoperative neurological impairments

J Neurosurg / Volume 119 / July 2013 11 Y. H. Kim et al.

TABLE 4: Risk factors for postoperative neurological deficits following SMA resection*

Multivariate Analysis Factor Univariate Analysis, p Value p Value OR (95% CI) patient sex 1.000 NI patient age (>30 yrs) 0.131 0.264 2.278 (0.537–9.666) Szs duration (<10 yrs) 0.069 0.108 3.391 (0.765–15.041) Sz frequency (<1/wk) 0.223 NI Sz type 0.758 NI predisposing factor 1.000 NI side of surgery 1.000 NI type of surgery 0.213 NI SMA proper resection 0.078 0.204 2.893 (0.561–14.922) pre-SMA resection 0.756 NI cingulate gyrus resection 0.023 0.027 6.530 (1.234–34.562) op time (>300 mins) 0.354 NI histological diagnosis (tumors) 0.206 NI unfavorable Sz outcome (Engel Classes III & IV) 0.203 NI

* Boldface indicates statistical significance. Abbreviation: NI = not included. following SMA resection. The cingulate gyrus is a major impairments associated with cingulate cortical resection medial cortical structure that overlies the corpus callo- may be induced by SMA damage. This phenomenon was sum from the lamina terminalis rostrally to the splenium also found in a recent series described by von Lehe et at its caudal extent.15,21 Traditionally, the cingulate gyrus al.25 In their 22 resective surgeries for cingulate gyrus has been associated with limbic and emotion pathways.24 epilepsy, extracingulate resection, including the SMA, However, recent neurophysiological and neuroanatomical induced SMA syndrome. The aforementioned studies21,25 studies of humans and animals have dramatically changed and the present study demonstrate that resection of both our understanding of cingulate function.15 Specifically, the the SMA and the cingulate gyrus induces postoperative cingulate cortex has been recognized to consist of 4 divi- neurological deficits. Consequently, SMA resection, when sions, each with a unique function: the perigenual anterior accompanied by resection of the cingulate cortex below cingulate cortex (pACC), the midcingulate cortex (MCC), it, is likely to cause postoperative neurological deficits, the posterior cingulate cortex (PCC), and the retrosplenial although the exact mechanism of this and the connections cortex (RSC).24 Among these cortices, the MCC consti- between these 2 areas require further study. tutes approximately one-third of the total cingulate cortex In the present study, 3 patients (7.0%) experienced beneath the SMA, and it functions to 1) select an appropri- permanent Grade 4 motor weakness. All of these pa- ate motor response based on its motivational significance, tients underwent resection of the most posterior part of and 2) carry out that response via direct connections to the the SMA without definite cortical damage or resection of spinal cord, limbic cortices, and primary or supplemen- the precentral gyrus. Moreover, these patients were prone tary motor areas.13,21 The other cortices—the pACC, PCC, to postoperative motor impairments. Presumably, any un- and RSC—are associated with emotional, visuospatial, or noticed white matter injury of the primary motor fibers, memory functions rather than motor functions.3,21,22,24 Re- such as the corticospinal tract, could cause permanent mo- sected cingulate cortical areas corresponded to the MCC tor weakness. Peraud et al.17 and Kasasbeh et al.7 showed beneath the SMA in this study. Moreover, the 12 patients that the distance between the resected area and the pre- who underwent resection of both the SMA and the cin- central gyrus was significantly related to both transient gulate gyrus experienced postoperative motor weakness or and permanent neurological impairments. Moreover, the apraxia. The resected area of the cingulate gyrus may thus anterior-posterior extent of the resected SMA was the be responsible for motor functions, because resecting this most frequently reported risk factor of SMA syndrome in area produces motor impairments. previous reports.5,23,26 Resection of a longer anterior-pos- In our opinion, cingulate gyrus resection has not terior extent of the SMA may have caused white matter been studied adequately. Tate et al.21 reported on 90 re- injuries that went unnoticed, while sparing the primary sections of cingulate gyrus gliomas, the largest surgical motor area of any damage. However, cingulate gyrus re- series to date. In their study, a high rate of postoperative section was not associated with permanent neurological morbidity, such as SMA syndrome, was induced by resec- risk after SMA resection. Consequently, cingulate gyrus tion of the MCC or PCC when compared with resection resection with SMA resection was the significant risk fac- of the pACC. The authors demonstrated that these mor- tor for SMA syndrome, which induces mostly transient bidities occurred only in cases with SMA resection as a motor weakness or language problems. surgical trajectory. Therefore, postoperative neurological In this study we evaluated the risk factors for postoper-

12 J Neurosurg / Volume 119 / July 2013 Neurological deficits after SMA resection

Fig. 2. Case 35. Intraoperative photographs and postoperative brain MR images. The patient underwent implantation of a subdural grid to localize the ictal onset zone (A) as well as resection of both the right SMA and right cingulate gyrus (B). White lines indicate the right precentral sulcus. Postoperative axial (C), sagittal (D), and coronal (E) MR images depicting the resected area. The patient experienced mild left-sided weakness and motor apraxia for 8 days following the surgery. ative neurological deficits after SMA resection. Moreover, Disclosure the significance of SMA resection accompanied by cingu- This research was supported by the Original Technology late gyrus resection was discussed for the first time in the Research Program for Brain Science through the National Research literature. However, even this large analysis is not without Foundation of Korea funded by the Ministry of Education, Science limitations. The inclusion of heterogeneous disease groups and Technology (Grant No. 2011-0030103). and the retrospective nature of the study may have affected Author contributions to the study and manuscript preparation the results directly or indirectly. In particular, the heteroge- include the following. Conception and design: Chung, YH Kim, neous pathology of this study was a significant limitation. Lee, Han, CY Kim. Acquisition of data: YH Kim. Analysis and interpretation of data: Chung, YH Kim, CH Kim, JS Kim. Drafting Further analyses including more patients are necessary to the article: YH Kim. Critically revising the article: all authors. determine the relationship between resection of the most Reviewed submitted version of manuscript: all authors. Approved posterior aspect of the SMA and permanent neurological the final version of the manuscript on behalf of all authors: Chung. deficits. Potentially unnoticed white matter injuries should Statistical analysis: YH Kim, CH Kim, JS Kim. Administrative/tech- be assessed using diffusion tensor imaging. In addition, nical/material support: Chung, CH Kim, Lee, Han, CY Kim. Study topographical analysis of the cingulate gyrus is necessary supervision: Chung, YH Kim, Han, CY Kim. to define the differences between the motor functions of the ventral and dorsal cingulate. References 1. Bannur U, Rajshekhar V: Post operative supplementary motor area syndrome: clinical features and outcome. Br J Neurosurg Conclusions 14:204–210, 2000 After SMA resection, 47% of patients experienced 2. Bleasel A, Comair Y, Lüders HO: Surgical ablations of the mesial frontal lobe in humans. Adv Neurol 70:217–235, 1996 transient neurological symptoms, and 7% experienced 3. Chassagnon S, Minotti L, Kremer S, Hoffmann D, Kahane P: permanent deficits. Cingulate gyrus resection that was ac- Somatosensory, motor, and reaching/grasping responses to di- companied by SMA resection was significantly associat- rect electrical stimulation of the human cingulate motor areas. ed with postoperative transient neurological impairments. J Neurosurg 109:593–604, 2008

J Neurosurg / Volume 119 / July 2013 13 Y. H. Kim et al.

4. Duffau H, Lopes M, Denvil D, Capelle L: Delayed onset of the 17. Peraud A, Meschede M, Eisner W, Ilmberger J, Reulen HJ: supplementary motor area syndrome after surgical resection of Surgical resection of grade II astrocytomas in the superior the mesial frontal lobe: a time course study using intraoperative frontal gyrus. Neurosurgery 50:966 –977, 2002 mapping in an awake patient. Stereotact Funct Neurosurg 18. Rosenberg K, Nossek E, Liebling R, Fried I, Shapira-Lichter 76:74–82, 2001 I, Hendler T, et al: Prediction of neurological deficits and re- 5. Fontaine D, Capelle L, Duffau H: Somatotopy of the supple- covery after surgery in the supplementary motor area: a pro- mentary motor area: evidence from correlation of the extent spective study in 26 patients. Clinical article. J Neurosurg of surgical resection with the clinical patterns of deficit. Neu- 113:1152–1163, 2010 rosurgery 50:297–305, 2002 19. Rostomily RC, Berger MS, Ojemann GA, Lettich E: Postopera- 6. Hatipoğlu MA, Prabhu S, Weinberg J, Shah K, Sawaya R: tive deficits and functional recovery following removal of tu- Surgical treatment of supplementary motor area lesions. Turk mors involving the dominant hemisphere supplementary motor Neurosurg 19:306–307, 2009 area. J Neurosurg 75:62–68, 1991 7. Kasasbeh AS, Yarbrough CK, Limbrick DD, Steger-May K, 20. Russell SM, Kelly PJ: Incidence and clinical evolution of Leach JL, Mangano FT, et al: Characterization of the sup- postoperative deficits after volumetric stereotactic resection plementary motor area syndrome and seizure outcome after of glial neoplasms involving the supplementary motor area. medial frontal lobe resections in pediatric epilepsy surgery. Neurosurgery 52:506–516, 2003 Neurosurgery 70:1152–1168, 2012 21. Tate MC, Kim CY, Chang EF, Polley MY, Berger MS: As- 8. Kim YH, Kim CH, Kim JS, Lee SK, Chung CK: Resection sessment of morbidity following resection of cingulate gyrus frequency map after awake resective surgery for non-lesional gliomas. Clinical article. J Neurosurg 114:640–647, 2011 neocortical epilepsy involving eloquent areas. Acta Neuro- 22. Turken AU, Swick D: Response selection in the human ante- chir (Wien) 153:1739–1749, 2011 rior cingulate cortex. Nat Neurosci 2:920–924, 1999 9. Krainik A, Duffau H, Capelle L, Cornu P, Boch AL, Man- 23. Ulu MO, Tanriöver N, Ozlen F, Sanus GZ, Tanriverdi T, Oz- gin JF, et al: Role of the healthy hemisphere in recovery after kara C, et al: Surgical treatment of lesions involving the sup- resection of the supplementary motor area. Neurology 62: plementary motor area: clinical results of 12 patients. Turk 1323–1332, 2004 Neurosurg 18:286–293, 2008 10. Krainik A, Lehéricy S, Duffau H, Capelle L, Chainay H, Cor- 24. Vogt BA: Pain and emotion interactions in subregions of the nu P, et al: Postoperative speech disorder after medial frontal cingulate gyrus. Nat Rev Neurosci 6:533–544, 2005 surgery: role of the supplementary motor area. Neurology 60: 25. von Lehe M, Wagner J, Wellmer J, Clusmann H, Kral T: Epi- 587–594, 2003 lepsy surgery of the cingulate gyrus and the frontomesial cor- 11. Laplane D, Talairach J, Meininger V, Bancaud J, Orgogozo JM: tex. Neurosurgery 70:900–910, 2012 Clinical consequences of corticectomies involving the supple- 26. Zentner J, Hufnagel A, Pechstein U, Wolf HK, Schramm J: mentary motor area in man. J Neurol Sci 34:301–314, 1977 Functional results after resective procedures involving the 12. Lim SH, Dinner DS, Pillay PK, Lüders H, Morris HH, Klem supplementary motor area. J Neurosurg 85:542–549, 1996 G, et al: Functional anatomy of the human supplementary sen- 27. Zilles K, Schlaug G, Matelli M, Luppino G, Schleicher A, Qü sorimotor area: results of extraoperative electrical stimulation. M, et al: Mapping of human and macaque sensorimotor areas Electroencephalogr Clin Neurophysiol 91:179–193, 1994 by integrating architectonic, transmitter receptor, MRI and 13. Morecraft RJ, Van Hoesen GW: Cingulate input to the pri- PET data. J Anat 187:515–537, 1995 mary and supplementary motor cortices in the rhesus mon- key: evidence for somatotopy in areas 24c and 23c. J Comp Neurol 322:471–489, 1992 14. Nelson L, Lapsiwala S, Haughton VM, Noyes J, Sadrzadeh AH, Moritz CH, et al: Preoperative mapping of the supple- Manuscript submitted July 31, 2012. mentary motor area in patients harboring tumors in the me- Accepted March 25, 2013. dial frontal lobe. J Neurosurg 97:1108–1114, 2002 Please include this information when citing this paper: pub- 15. Paxinos G, Mai JK: The Human Nervous System, ed 2. San lished online May 3, 2013; DOI: 10.3171/2013.3.JNS121492. Diego: Elsevier Academic Press, 2004 Address correspondence to: Chun Kee Chung, M.D., Ph.D., De 16. Penfield W, Welch K: The supplementary motor area of the ce- partment of Neurosurgery, Seoul National University College of rebral cortex; a clinical and experimental study. AMA Arch Medicine, 101 Daehangno, Jongno-gu, Seoul 110-744, Korea. email: Neurol Psychiatry 66:289–317, 1951 [email protected].

14 J Neurosurg / Volume 119 / July 2013